Spin Welding Plastic: Process, Joint Design & Defects

A production engineer watches a batch of filled chemical drums fail a leak test, and the culprit is not the moulding, the gasket or the cap thread. It is a half-second of hesitation in how the cap was spun against the drum neck. That tiny friction weld is the whole seal, and when it works, nobody notices it at all.

Spin Welding Plastic: The Friction Process That Seals in Seconds

On a filling line, a production engineer watches a batch of filled chemical drums fail a leak test. The moulding looks clean, the gasket is seated, and the cap thread is within tolerance. The problem is a half-second of hesitation in how the cap was spun onto the drum neck. That tiny friction weld is the entire seal. When the process is right, nobody notices it; when it drifts, the leak test finds it immediately.

Spin welding, also called rotary friction welding, joins two thermoplastic parts with rotationally symmetrical joint faces. One half is held in a fixture while the other is spun and pressed against it. Friction at the contact surface melts the polymer, rotation stops, and pressure is maintained while the melt consolidates into a joint. The joint face can be circular, annular or conical, according to the scope of DVS technical code 2218-1. That geometric limit is important: spin welding is not a universal plastic joining method.

Its appeal is easy to understand. The process needs no adhesive, no consumable filler, and no external heat source. Cycle times are often measured in seconds, and the result can be a genuinely hermetic seal. It is a production process rather than a repair trick, which is why it appears on drum caps, filter housings, floats, reservoirs and medical components.

It is not ultrasonic welding, which uses high-frequency vibration from a horn and suits many non-round parts. It is not hot plate welding, which melts parts against a heated platen. It is not a mechanical fastener, and it cannot join thermosets, because those materials do not melt and reflow. Spin welding depends on a thermoplastic’s ability to soften, flow and re-solidify.

The Vocabulary You Will Meet Later

The weld interface is the contact area where melting and bonding occur. Flash is the molten polymer squeezed out of that interface. Collapse, also called upset, is the shortening of the parts as material is displaced; the related term weld depth describes how far the mating faces move together. Hold time is the period after rotation stops when pressure keeps the joint closed during cooling. A drive feature is a flat, lug, spline or knurl that lets the spindle grip the rotating part, while the fixture locates and supports the stationary half.

Where the Heat Comes From

There is no hidden heater. Mechanical work at the interface becomes thermal energy: friction force multiplied by surface speed, applied over time. The polymer does not need to reach a bulk melting temperature throughout the part. It needs a thin interfacial layer to melt and flow while the surrounding material stays firm enough to contain it.

TWI describes four phases in thermoplastic friction welding: contact under force with no material flow, onset of melting and displacement, a steady state, and cooling under maintained force. Phases I and II typically take about 0.5 to 8 seconds depending on joint area, while cooling commonly takes 4 to 10 seconds. Weld and cooling pressures usually fall between 0.5 and 2 MPa.

The governing physical requirement is simple: heat must be generated faster than the polymer conducts it away. If conduction wins, the interface never develops a proper melt layer. If friction wins by too much, the melt becomes excessively fluid, gets expelled as flash, and the joint can end up starved. Every parameter discussion later in this article comes back to that balance.

The idea is older than engineering plastics themselves. Dukane notes that spin welding of metals has been practised for at least a hundred years, and that it is one of the oldest methods of joining thermoplastic parts. Metal rotary friction welding proved the principle; polymer grades and machine controls made it a packaging and assembly staple.

The rest of this guide follows the production problem from beginning to end: how the cycle works, which resins can take the heat, how to design the joint, how to set RPM and pressure, how spin welding compares with other methods, where it pays off, how to troubleshoot defects, what safety controls are needed, and how to move from a first prototype to a stable process. If you also work in metal fabrication, some of the discipline will feel familiar; process windows matter as much as they do in aluminum MIG welding.

How the Spin Welding Cycle Actually Works, Step by Step

Once a joint has been designed for rotation, the bond itself is made in a matter of seconds. What looks like a simple spin and press from the outside is actually a tightly controlled sequence of motion, friction, and consolidation. Each step has to hand off cleanly to the next. A fraction of a second out of place can be the difference between a hermetic seal and a scrap part.

The Full Production Cycle

A complete spin welding cycle follows this sequence in production:

1. Part loading and fixturing: The two halves of the assembly are placed into their dedicated fixtures. The stationary half is clamped down, while the rotating half is secured to the spindle. Fixturing decides whether the weld faces meet flat and concentric. Runout, clamp rigidity, and axial alignment all matter here, and the precision required is no different from the tight tolerances seen in aerospace TIG welding.
2. Drive engagement and run-up: The spindle is brought up to its target rotational speed. On some machines this happens under little or no axial load, while others begin to apply pressure as the part approaches speed. The aim is a repeatable surface speed at the joint, not simply the highest RPM available.
3. Friction phase: The rotating part is brought into contact with the stationary part under axial pressure. At that moment, friction begins converting mechanical energy into heat at the weld interface. Torque rises, the polymer softens, and the machine transitions from cold contact to a working melt film.
4. Melt and steady-state displacement: The interfacial polymer reaches its melt temperature and begins to flow. The two parts are forced together, causing a controlled shortening of the assembly, often called weld collapse. On a well-instrumented machine, that displacement is tracked in real time because it shows whether the melt is forming as expected.
5. Braking/stop: Rotation is halted as quickly and deliberately as possible to prevent any relative motion while the polymer is still viscous. The stop needs to be fast, but not so violent that it shocks the fixture or twists the softening joint.
6. Hold/cool: Axial pressure is maintained with no rotation, allowing the molten polymer to consolidate and re-solidify under load. This phase shapes the final weld and helps the joint develop strength before the clamp opens.
7. Unload: The clamp is released and the now fully bonded part is removed, ready for inspection. Operators typically check flash formation, weld collapse, concentricity, and any signs of incomplete bonding.

What the Machine Is Adjusting in Those Seconds

Although the sequence sounds simple, the machine is balancing several variables at once: rotational speed, axial force, contact time or displacement, and brake and hold timing. Change one and the heat balance shifts. More speed raises the surface velocity at the joint. More force generates heat faster and affects how much molten polymer remains in the interface. Time and displacement determine when the machine stops adding energy.

That is why production setups are built around a proven recipe and then verified with destructive and non-destructive checks, rather than judged by sight alone. The cycle may last only seconds, but every stage leaves a signature on the finished weld.

Which Plastics Can Be Spin Welded — and Which Cannot

When it comes to material selection for spin welding plastic, the first rule is simple: both halves must be thermoplastics. Beyond that, the key is compatibility. According to Dukane, the two parts should be chemically compatible, have similar melt temperatures, and have similar melt flow indices. If any of these fall too far apart, a consistent bond becomes incredibly difficult to achieve.

The Science: Friction vs. Heat Loss

Not every thermoplastic will generate enough heat at the weld interface to form a true melt. The deciding factor is the ratio of a resin’s coefficient of friction to its thermal conductivity. As Dukane states, virtually all thermoplastics can be spin welded if this ratio is high enough that heat is created faster than it dissipates away from the joint. If a material is too thermally conductive, it will wick the frictional heat into the surrounding walls before melting can occur.

This heat generation behavior is what splits plastics into two distinct groups: amorphous and semi-crystalline.

Amorphous vs. Semi-Crystalline Resins

Amorphous resins, such as polycarbonate, polystyrene, ABS, and rigid PVC, soften gradually over a wide temperature range. Rather than hitting a sharp melting point, they transition into a rubbery, weldable state with relative ease. This broad softening window makes them far more forgiving to slight fluctuations in RPM, pressure, or part temperature.

Semi-crystalline resins, such as polyethylene, acetal, nylon, PET, and PEEK, behave the opposite way. They remain rigid up until their very distinct melt point, and once reached, require a sharper burst of higher energy input to get there. Even more critically, they re-solidify incredibly fast after rotation stops. This narrow window leaves far less room for error, especially when working with tight cycle times. Supporting this, DVS 2218-2 notes that while PE and PP are very well suited to rotational friction welding, cross-linked polyolefins range from poorly weldable to not weldable at all.

Practical Weldability Ratings

For a quick reference, Branson’s joint design guide rates weldability on a scale of 1 (easiest) to 5 (most difficult). With this scale in mind, the following are general expectations:

• 1 – 2 (Easiest): ABS, polycarbonate, rigid PVC, and polystyrene.
• 2 – 3 (Moderate): Acetal and PET.
• 3 (Moderate-Difficult): Polyethylene.
• 3 (More Demanding): PEEK.
• 3 – 4 (Hardest): Fluoropolymers.

Fillers, Moisture, and Contaminants

Filled and reinforced grades are a common point of confusion. As Dukane notes, when materials are reinforced with glass, the reinforcement provides no benefit at the joint. The friction and melt flow encourages glass fibres to lie parallel to the weld line rather than crossing it. Not only is strength reduced, but fibre orientation can also create warpage and uneven flash. For this reason, you should never rely on the filled grade’s tensile data to predict joint strength.

Moisture is another dealbreaker for hygroscopic resins. Nylon, polycarbonate, and PET must be properly dried prior to moulding and welding. Any retained moisture will flash to steam at the melt interface, creating porous, weak welds.

Other variables can also sabotage a bond. Dissimilar-material pairs should only be attempted if chemically compatible. Excessive regrind, colourants, lubricants, mould release agents, or certain flame retardants can all act as a contamination layer, preventing proper fusion.

Comparison of Common Resins

Resin	Type	Branson Ease (1-5)	Practical Notes
ABS	Amorphous	1 – 2	Very forgiving, great for general enclosures.
Polycarbonate (PC)	Amorphous	1 – 2	Easy to weld, but must be dried to avoid porosity.
Rigid PVC	Amorphous	1 – 2	Weldable, but requires strict fume control due to decomposition risk.
Polystyrene (PS)	Amorphous	1 – 2	Forgiving, but can be brittle at the joint.
Polyethylene (PE)	Semi-crystalline	3	Very well suited per DVS 2218-2, but needs precise energy control.
Acetal (POM)	Semi-crystalline	2 – 3	Good candidate, but can degrade if overheated.
PET	Semi-crystalline	2 – 3	Requires thorough drying and tight process control.
PEEK	Semi-crystalline	3	Possible, but demands high, consistent energy input.
Fluoropolymers	Varies	3 – 4	Hardest to weld; requires extensive testing and strict fume precautions.

Joint Design for Spin Welding: Butt, Tongue-and-Groove and Shear Joints

A spin weld is won or lost at the drawing stage. The joint has two jobs at once: generate enough friction to melt polymer, then contain that melt while it consolidates. A geometry that makes heat but gives the melt nowhere to go produces flash, voids and a weak seam. A geometry that contains melt but never reaches the right interfacial temperature produces a cold weld that may pass a visual check and fail in service.

Good design balances those jobs. The weld area must be large enough to transmit torque and absorb axial load, the mating faces must align before they touch, and any displaced material needs a deliberate place to go. Branson’s joint design guidance frames the basics well: a weld area larger than a typical wall section, part-to-part alignment features, and a drive feature on the spinning half.

The Main Joint Configurations

• Flat butt joint: Two flat annular faces rub together. It is simple and cheap to tool, but weld area is limited and flash escapes outward as a visible ring. Best for low-pressure parts where cosmetics matter less than cost.
• Step or flanged shear joint: A stepped vertical wall creates a long shear path. This is the strongest and most reliable option for pressure and hermetic parts, because melt is contained along the step rather than squeezed sideways.
• Tongue-and-groove with flash trap: The tongue enters a groove and displaced melt is captured in a recess. It self-locates the parts and hides flash, making it a strong choice for appliance, filter and medical housings.
• Tongue-and-groove with external skirt: A skirt outside the joint protects the weld, hides flash and adds stiffness. Useful where the joint is visible or exposed to handling damage.
• Return flange: A folded flange adds weld area and stiffness to thin walls without thickening the whole part. It suits containers and shells where nominal wall stock is limited.

Dimensions That Decide Success

Interference is the first number to get right. Extol specifies 0.25 to 0.50 mm (0.010 to 0.020 in) for joint interference, while Branson lists a similar 0.010 to 0.020 in range depending on part size and tolerance. Too little interference and the faces polish each other without melting. Too much and torque, flash and distortion climb quickly.

Add a lead-in chamfer of 0.020 to 0.040 in minimum, as Branson recommends, so the parts guide together instead of catching on a sharp edge. Generous root radii reduce stress concentration. The weld area should exceed the typical wall section, and the flash trap cross-section should be greater than the area of the displaced weld rib, per Extol, or the trap fills and melt is forced somewhere unwanted. Walls must also resist the weld force; Extol notes that parts should handle roughly 50 to 400 pounds.

Torque, Centering and Angular Position

The spinning half needs a drive feature: flats, lugs, splines or a knurled band that transmits torque without slipping. Clamp and drive only on surfaces where marks are acceptable. Equally important, no contact should occur outside the intended joint area. A rib rubbing on a fixture or a skirt dragging on the mating part steals energy and generates debris.

Concentricity matters because an offset joint melts unevenly, producing a thick flash line on one side and an unbonded sector on the other. Out-of-round mouldings make this worse. Check runout before blaming the machine. Also remember that spin welding leaves a random final angle. If the cap, port or label must index, add a hard stop or anti-rotation feature rather than expecting the process to clock itself.

Keeping Flash and Particles Out of the Product

Dukane notes that spin welding creates particulate because softened surface material is stripped and rolled into small balls before true melting begins. For fluid or medical paths, design an internal trap that captures this material away from the flow, with smooth radii and no dead pockets. Clean mating faces before trials; the same discipline used to remove oils before metal welding applies here, as covered in our aluminum cleaner for welding guide.

Joint type	Strength	Flash control	Typical use
Flat butt	Low to medium	Poor; visible ring	Low-pressure covers, toys
Step or flanged shear	High	Good with trap	Pressure vessels, fuel parts
Tongue-and-groove + trap	Medium to high	Excellent	Filters, appliances, medical
External skirt	Medium	Good; hidden	Visible housings
Return flange	Medium	Fair	Thin-wall containers

Setting the Parameters: RPM, Pressure, Weld Depth and Hold Time

Joint design tells you where the heat must go. Machine settings decide whether enough heat arrives there, fast enough, before the polymer conducts it away. The tricky part is that the numbers on the HMI are not universal. A program that seals a 20 mm medical connector will scorch a 200 mm drum cap. Parameters have to be derived from the joint in front of you.

Start With Surface Speed, Not a Borrowed RPM

Friction heat depends on how fast the two surfaces slide past each other, not on spindle RPM alone. A small part spinning at 10,000 rpm may have a lower surface speed than a large part at 2,000 rpm. That is why copying a program from another cell is such a common and expensive mistake.

Dukane’s spin welding guidance offers a useful starting rule: target RPM is roughly 8,000 divided by the joint diameter in inches, or 200,000 divided by the diameter in millimetres. For a 60 mm cap, that gives 200,000 ÷ 60 ≈ 3,330 rpm. Dukane’s primer notes that successful applications can deviate by 50 percent or more. Spin welding literature generally cites speeds from about 200 to 14,000 rpm, with roughly 2,000 rpm often quoted as a normal working speed. Treat the formula as a first guess, not a recipe.

Why the spread? Resin chemistry, joint diameter, wall stiffness, melt viscosity and pressure profile all shift the sweet spot. The formula gets you into the right neighbourhood. Trials tell you which house on the street.

Pressure: Secondary Parameter, Primary Consequences

Pressure generates friction at the start and consolidates melt at the end. Thermoplastic friction welding pressures generally sit between about 72.5 psi and 290 psi, or 0.5 to 2 MPa. Below that band, heat builds slowly, cycles stretch out, and the interface may never reach a proper melt. Above it, melt gets squeezed out before it can do its job. TWI’s guidance on friction welding of plastics is blunt: excessive pressure can force out most of the molten thermoplastic and produce a cold weld, even when the parts look joined.

Production machines therefore use multi-stage profiles. A higher friction pressure brings the interface up to temperature quickly. A lower weld or consolidation pressure then holds the melt in place while the joint forms. Some recipes add a brief transition stage to avoid a pressure spike at the stop. The numbers are resin-specific; the principle is not.

Time, Displacement and the Hold That Makes the Joint

Weld time can be controlled by the clock or by collapse. A time trigger is simple, but it assumes every part starts at the same height and heats at the same rate. It suits tightly held mouldings and simple joints. Displacement control watches how far the parts have collapsed and switches stages when that target is reached. When flange flatness or moulding height varies, displacement is usually the more repeatable master.

The displacement trigger is also your weld depth: the amount of material consumed at the interface. Set it from the joint’s interference and flash-trap volume, then verify by measuring actual collapse on production parts. Once steady-state displacement is reached, TWI notes there is generally no strength benefit in prolonging that phase. Extra time adds flash, heat and cycle cost without making the weld stronger.

Hold, or cool, time is different. After rotation stops, the interface is still molten and weak. Clamp force must stay on until the polymer has cooled enough to carry load. Release early and the joint can peel, crack or relax open. In semi-crystalline resins, the hold also influences recrystallisation and residual stress.

Finding the Window, Not a Magic Number

Run a small matrix, not a single trial. Change RPM first, then pressure, then the displacement trigger, measuring weld strength and collapse at each point. Cross-section the best and worst samples. The goal is a robust middle where normal variation in parts, ambient temperature, humidity, part temperature, machine repeatability and fixture wear still produces a good weld. A setting that only works at the edge of the window is not a production setting.

The logic mirrors what makes an aluminum MIG welding setup repeatable: establish a window, then protect it from drift.

Starting points by joint diameter, not specifications

Joint diameter	Formula RPM	Trial band (±50%)	Practical note
20–40 mm	5,000–10,000	2,500–15,000	Fast cycles; watch flash
40–80 mm	2,500–5,000	1,250–7,500	Use a pressure profile; displacement stop
80–150 mm	1,300–2,500	650–3,750	Stiff fixture; check ovality
150–300 mm	670–1,300	335–1,950	High clamp stiffness; verify collapse

Across all sizes, begin pressure near the middle of the 0.5–2 MPa band and adjust from there.

Note: Document approved settings by material lot, colourant and regrind percentage, not just part number. When caps start failing, the first question is “what changed in the resin?” The settings log is how you answer it.

Spin Welding vs Ultrasonic, Vibration, Hot Plate and Laser Welding

There is no “best” way to weld plastic. There is only the process that fits your part geometry, your resin, your seal requirement and your budget. Shops that get into trouble usually picked a method first and designed the part around it afterwards. So before committing to spin welding, it helps to see it standing next to its competitors, with its weaknesses showing.

Six things separate the processes in practice: whether the joint is round or arbitrary, how big the part is, the resin, whether you need a hermetic seal, how much flash and particulate you can tolerate, and the tooling and capital cost per joint.

The Geometric Divide: Round vs Non-Round

Spin welding requires a circular joint axis. Full stop. That single constraint is the clearest dividing line between it and ultrasonic or vibration welding, which can follow non-round joint lines around complex outlines. If your housing is rectangular, spin welding is simply not on the table, and no amount of process development will change that.

Ultrasonic Welding

Ultrasonic welding shines on small, complex, non-round parts, and it is fast and easily automated. The cost is in the tooling: each part needs a horn shaped to its geometry, and that horn wears. Energy delivery also drops off over large or very soft, highly amorphous parts, and the high-frequency vibration at the joint can shed particulate. Joint design is unforgiving, with energy directors doing most of the work.

Vibration and Orbital Welding

Linear vibration handles rectangular and large parts and produces strong, repeatable seals, while orbital welding covers similar ground with a different motion path. Both carry high capital cost and substantial fixtures, and both need rigid, well-supported walls. For a big rectangular battery tray or a dashboard duct, they are often the only realistic answer.

Hot Plate Welding

Hot plate welding is beautifully simple: heat both faces on a platen, withdraw it, press together. It suits large and irregular parts and is a mainstay for polyolefins such as PE and PP, which spin weld less easily. Its weaknesses are cycle time, stringing as the platen withdraws, and dwell control. Material choice matters too — as TWI notes, PTFE-coated hot plates running above 270 °C can cause polymer fume fever, so fluoropolymers demand extra care.

Laser and Infrared Welding

Laser and infrared welding are the clean options: minimal flash, very low particulate, tight control, and the ability to follow a contour with a scanning head. They are the default choice for medical devices and sensitive electronics housings. The trade-off is equipment cost and the need for a transmissive top layer or a precisely matched absorbent interface. If you are weighing laser systems generally, this guide to Alpha laser welding systems and safety covers how the hardware side works.

Solvent and Adhesive Bonding

No equipment, no capital, and sometimes a genuinely strong joint on resins like ABS or acrylic. But you inherit cure time, VOC exposure, surface preparation, and consistency that depends heavily on the operator. For a pressure-tight, high-volume part, bonding is usually the wrong tool.

Where Spin Welding Wins — and Where It Doesn’t

Spin welding wins on round, medium-to-large parts that need a strong hermetic seal at high throughput. Dukane describes it as a fast way to join circular joints and very reliable at delivering a hermetic seal, and equipment makers also note it is aggressive enough to weld materials that sonic welding struggles with. Standards work such as DVS 2218-2 for PE and PP mouldings reflects how mature the process is, including machines offering constant or variable speed with constant or variable weld force.

It loses on square parts, delicate internal geometry that cannot survive rotation, and any assembly where the final angular position must be exact — spin welding leaves a random index angle unless you add a hard stop.

Process	Geometry	Best Fit	Watch Out For
Spin	Round only	Medium–large, hermetic, high volume	Flash, random angular index
Ultrasonic	Any, small	Complex small parts, electronics	Horn cost and wear, particulate
Vibration / orbital	Rectangular, large	Big housings, strong seals	High capital, rigid walls needed
Hot plate	Any, large	PE/PP, low volume, big parts	Slow, stringing, fume risk
Laser / IR	Any, contour	Clean medical, low particulate	Equipment cost, layer matching
Solvent / adhesive	Any	Prototypes, low volume	Cure time, VOCs, variability

Benefits, Limitations and the Applications Where Spin Welding Pays Off

Spin welding is easy to admire on a shop floor and surprisingly hard to justify on a spreadsheet until you look at the right numbers. The case is not that it is the cheapest way to join plastic. The case is that on certain round parts, it delivers a sealed, strong joint in a couple of seconds with almost no consumable spend, and that combination is very difficult to beat with anything else.

Benefits That Show Up on the Balance Sheet

• No consumables and no adhesives. No solvent drums, no primer, no shelf-life management, no cure ovens, no VOC paperwork for the bonding area.
• Fast cycles. Wikipedia’s process description notes that parts can typically be loaded by hand in 2 to 5 seconds and press actuation takes 1 to 2 seconds in a production setting, which puts many joints comfortably under ten seconds total.
• Strong joints. Forward Tech states that spin welding often yields a weld strength consistently equal to or stronger than other areas of the part, so the joint can be designed to carry similar stresses to the surrounding wall.
• Hermetic seals are realistic. A continuous circumferential weld with no gasket is the reason filled drums, medical housings and filter bodies choose this process.
• Wide size and material range. Equipment makers note the process handles both large and small diameter parts as well as a broad range of thermoplastic resins, from small caps to tank closures.
• Simple tooling and easy automation. Fixtures are usually machined blocks and drive features, not custom ultrasonic horns, and the motion is a spindle plus a press, which is straightforward to robot-tend.

Limitations You Have to Design Around

• Round joints only. Rectangular or irregular joint lines belong to vibration, hot plate or ultrasonic welding.
• One part must rotate. Heavy, unbalanced or pre-filled assemblies are awkward or impossible to spin safely.
• Flash and particulate are inherent. Melt is displaced by definition, so containment is a design task, not an option.
• Angular orientation is uncontrolled. If a cap must index to a label or a port must face a fixed direction, you need a hard stop or a post-weld orienting feature.
• Parts get shorter. Spin welding produces axial collapse, so every drawing needs a stack-up allowance for the new overall length after welding.
• Glass-filled joints lose their reinforcement. Fibres do not cross the weld line, so the joint behaves more like the base resin than the filled grade.
• Real process development is required. Settings cannot be copied from a datasheet; they have to be proven on the actual parts.

Where It Earns Its Keep

Application family	Why spin welding fits
Automotive fluid reservoirs, fuel and vapour canisters	Round bodies, sealed joints, high volume, no adhesive compatibility issues
Filter housings, HVAC actuators, appliance pump bodies	Pressure-tight circumferential joint with simple machined fixtures
Plumbing fittings, PE and PP tanks, floats	Large diameters, polyolefins that weld well by friction
Medical and IV components	Clean internal path, no solvent residue, flash traps keep debris out of the lumen
Toy and consumer shells, industrial drums and closures	Fast cosmetic or hermetic closure at low cost per unit

The Economics: Cost Per Joint at Volume

Spin welding’s capital cost sits in a middle band. It is usually more than a simple heated-tool setup and less than a laser contour system, and its tooling is typically far cheaper than a set of ultrasonic horns. Where it wins is cost per joint at volume: short cycle time, no consumable line item, and very little scrap once the window is stable. Energy use per joint is modest because the machine only spins for a fraction of the cycle. The costs that bite are the ones people forget: prototype tooling iterations, destructive audit samples, and the engineering hours spent finding the parameter window.

No single joining method dominates everywhere. Even in metalwork, processes divide the territory by geometry and thickness; aluminum MIG welding owns one corner of that map and stick or TIG own others. Plastics behave the same way, and the shops that get the best results treat spin welding as a targeted tool rather than a universal answer. For the underlying process science, TWI’s friction welding of plastics reference is a solid starting point.

Defects, Troubleshooting and Quality Control That Actually Catches Problems

Spin weld cycles often last only a second or two, so small process drift can slip by until a whole batch fails a leak or burst test. A solid quality plan isn’t about inspecting more parts. It’s about knowing what goes wrong, why, and which checks earn their place.

Common Defects and Their Fixes

Not every flaw is equal. Some are cosmetic; others destroy a hermetic seal. Here are the most common failure modes:

• Weak or cold welds: As noted by TWI, excessive weld pressure can force out molten thermoplastic and produce a cold weld. Low RPM, mould release contamination, or wrong surface speed make it worse. Fix: lower consolidation pressure, clean the joint, and match RPM to the right linear speed for the diameter.
• Incomplete weld or unbonded sector: Out-of-round parts, warped flanges, misalignment, or uneven clamping prevent full contact. Dukane notes that out-of-round parts generate excess flash at the locations that are out of round. Fix: improve concentricity, rework fixturing, and validate clamping force around the full circumference.
• Excessive flash and particulate: Too much interference, excessive collapse, worn tooling, or an oversized weld rib over-displace melt. Dukane also notes that spinning beyond the point of true melting doesn’t increase bond line thickness or strength, but does throw more flash and particulate. Fix: switch to displacement control, cut the collapse allowance, and inspect flash traps for wear.
• Part distortion or ovality: Thin walls are especially vulnerable. Extol warns that parts must be rigid enough to resist weld force of roughly 50 to 400 pounds to prevent deflection during the cycle. Excess clamp force compounds it. Fix: reduce clamp pressure, add support ribs, and allow more even cooling.
• Weld-line read-through, sink, or short shots: Uneven melt pools, poor flash containment, or too little energy create sink marks without enough material to bond the interface. Fix: revisit the joint geometry to contain melt, and confirm energy input without over-expelling it.
• Internal particulate contamination: Flash migrates into fluid paths when it isn’t contained, a real concern for parts used in aerospace welding applications or medical housings. Fix: use a fully enclosed internal flash trap and validate with clean-path inspection after cross-sectioning.
• Inconsistent strength part to part: Machine drift, regrind variation, moisture content, and temperature swings are the usual culprits. Semi-crystalline resins such as acetal or nylon re-solidify fast, and without controlled cooling or annealing that locks in residual stress. Fix: dry moisture-sensitive resins, cap regrind percentages, and track shop conditions against process data.

Quality Control: Inspect in Order of Rigour

A robust inspection plan escalates with risk. A visual check may be enough for a large run of caps; for pressure vessels or safety-critical parts, the proof level must be far higher.

1. Visual and dimensional checks: Look for excessive flash, flash intrusion, ovality, and part deflection. Pay extra attention to parts headed into precision assemblies, such as alloy wheel welding repair housings or sealed actuation bodies.
2. Collapse/displacement monitoring: The shop’s greatest asset. Track axial collapse per cycle against your established range. Any deviation is usually the first sign of drift, long before a destructive failure.
3. Burst and pressure decay testing: For hermetic or pressurized joints these validate the seal, not just the bond. Pressure decay is sensitive enough to catch micro-leaks that visual inspection misses.
4. Peel and tensile testing: Destructive coupons give quantified strength data. The usual goal is a joint that fails in the parent material, not at the weld interface.
5. Cross-sectioning and microscopy: Cutting through the weld reveals melt flow, voids, unbonded sectors, and internal flash intrusion. This is the gold standard for validating a new joint design or chasing an intermittent failure.
6. Dye penetrant and destructive audit sampling: Dye penetrant highlights surface cracks around stressed areas, while routine destructive audits keep the batch within tolerance over time.

For safety-critical parts where failure would be catastrophic, 100 percent proof testing is sometimes the only honest option: pressure decay or burst on every part. The cost of a field failure always outweighs the cost of testing.

Troubleshooting Table and Process Control

Symptom	Likely Cause	Recommended Fix
Passes visual, fails pressure test	Cold weld from excess pressure or low RPM	Lower consolidation pressure, verify surface speed
Flash on one side	Out-of-round or misaligned flange	Check runout, true the fixture, enlarge the trap
Intermittent mid-run failures	Machine drift, moisture, regrind variation	Audit logs, dry resin, cap regrind %, stabilize temperature
Cracked or oval part	Excess clamp force, residual stress	Reduce clamping, support the part, add annealing
Flash in fluid path	Poor containment or excess collapse	Add an internal flash trap, use displacement control

At the end of the day, defects aren’t proof of a flawed process, they’re feedback. Track the data, escalate inspection with risk, and never trade containment for convenience.

Safety, Fume, Noise and Standards for Plastic Welding

With any form of friction welding, the biggest hazard isn’t the flame, it’s the unseen by-product of heat. During spin welding, the intense friction at the joint can rapidly bring polymers up to melt, and if pushed too far, past their degradation point. This can generate fine polymer particulate and, depending on the resin, potentially harmful decomposition products. Keeping these risks under control starts with understanding how to contain them before they reach the operator.

Fume, Particulate and Thermal Degradation

Not all plastics behave the same when overheated. Resins such as PVC, acetal, fluoropolymers, and those with flame-retardant grades carry a much higher risk of degrading under excess frictional heat. According to HSE guidance PPS13 on controlling fume during plastics processing, respiratory sensitisers, irritants and carcinogens can be produced, with correct melt temperature and residence time being the key controls. The HSE even gives a stark example: one grade of acetal processed at 240 °C has a safe residence time of only 7 minutes, versus 20 minutes at 205 °C. For shops working with fluoropolymers, TWI warns that hot gas welding of materials such as PVDF, ECTFE, PFA and FEP can cause influenza-like polymer fume fever, and that PTFE-coated hot plates should not exceed 270 °C. These figures reinforce the need to stay within your resin’s processing window, rather than cranking up energy to compensate for a poor joint design.

From a respiratory standpoint, OSHA’s welding fume guidance notes that acute exposure can cause eye, nose and throat irritation, dizziness and nausea, whilst prolonged exposure may cause lung damage and cancer. The most effective way to combat this is source extraction. NIOSH field studies of welding controls found local exhaust effectiveness varied wildly, with ventilated controls reducing exposure by up to 83 percent in the best case, whilst a canopy hood was completely ineffective. A shop vac is not LEV, it simply recirculates fine particulate. For this reason, many facilities model their containment around the same principles seen within air gas welding, where extraction at source is always prioritised over general dilution alone.

Noise, Mechanical and Ergonomic Hazards

Fume isn’t the only concern. High-speed spindles, part resonance and the sudden braking of inertia machines can all create noise levels that exceed comfortable exposure. This is best mitigated through purpose-built machine enclosures, acoustic barriers and correctly rated hearing protection.

Mechanically, spin welders present a unique blend of risks. Operators must be aware of rotating parts, stored kinetic energy within inertia flywheels, entanglement hazards, the ejection of hot flash, and the need for robust guarding and interlocks. On top of this, freshly ejected flash and fixtures can be hot enough to cause burns. Poor housekeeping only adds to the danger, with loose pellets and brittle flash creating slip risks across the floor. For high-volume lines, repetitive part loading can also create ergonomic strain, making adjustable fixturing and operator rotation well worth considering.

Standards, Documentation and a Safety Checklist

Whilst plastic friction welding doesn’t fall under every metal welding code, there are still concrete standards to lean on. The DVS 2218 series is widely referenced for rotational friction welding, alongside applicable ISO and ASTM test methods for welded joints. In the US, the framework is often

Practical Playbook: From First Prototype to Stable Production

Everything in this guide comes down to one question: can you turn a drawing into a joint that seals in month three the way it did on the first prototype? Five stages get you there, and skipping any one of them is how drum caps end up in the leak-test bin.

Stage 1: Feasibility

Start at a desk, not a machine. Confirm both halves are thermoplastics with compatible chemistry, similar melt temperature and similar melt flow index. Gather SDS sheets and datasheets, then pick the joint type from geometry and pressure duty rather than from whatever tooling already sits in the shop.

Then build the cheapest thing that can prove the concept: a straight-sided coupon or a single-cavity prototype tool. Weld a handful of samples, pull them apart, and read the failure. If the break runs along the interface instead of through parent material, you have a material or joint-design problem, and no amount of parameter tuning will rescue it.

Stage 2: Parameter Development

Derive a starting RPM from joint diameter, because surface speed at the interface is what generates heat, so smaller parts turn faster than large ones, as Dukane’s spin welding primer explains. Set pressure inside the usual 0.5–2 MPa band, and choose displacement-based control so flatness variation does not shift your weld depth. Then run a small matrix: vary RPM, vary pressure, hold everything else.

Measure collapse and strength on every sample, and cross-section the best and the worst. TWI’s guidance is to end weld time when the steady-state phase is reached, which a displacement transducer can identify directly. Machines built to DVS 2218-2 already log speed, friction time, forces, weld path, insert speed and cooling times, so use that data instead of guessing.

Stage 3: Design Freeze

Lock the joint dimensions once the process window is known: interference, weld area, flash trap volume, chamfers and radii. Treat every published dimension as a starting point only, because Branson is explicit that joint design guide figures are guidelines and specific applications may need variation.

Add locating features for part-to-part alignment, and a return flange beside the weld joint so thin walls are properly supported, per Extol’s design guidelines. Confirm stack-up with the collapse allowance built in, because the finished assembly will be shorter than the sum of its parts, every time.

Stage 4: Production Readiness

Release proper tooling drawings, list fixture wear parts with replacement intervals, and run a capability study on real parts from the real tool. Write a control plan: what you inspect, how often, against which limits. Train operators on loading, fixturing and what a bad weld looks like, since structured training pays back faster than most capital spending, much as a good aluminium welding course builds repeatable technique in metal shops.

Then write the reaction plan for out-of-spec parts: who stops the line, what gets quarantined, who signs the release. A reaction plan nobody has read is decoration.

Stage 5: Continuous Improvement

Track first-pass yield, scrap rate and cycle time weekly. Re-audit whenever the resin lot, colourant or regrind percentage changes, because those three variables move weld strength more than most engineers expect. Change one thing at a time and keep the old settings recorded so you can go back.

One-Page Checklist

• Both resins confirmed compatible, SDS on file, drying specified
• Joint type selected; flash trap and drive features drawn
• Starting RPM calculated from joint diameter
• Pressure set within 0.5–2 MPa; multi-stage profile defined
• Displacement control chosen; collapse target and tolerance set
• Three-sample matrix welded and cross-sectioned
• Fixtures and wear parts listed with spares on the shelf
• Control plan, inspection frequency and reaction plan signed
• Operators trained and assessed
• First-pass yield and scrap tracked weekly

Conclusion: The Seal You Cannot See

Go back to that drum line. When the cap weld is right, nobody on the floor thinks about it: the drum passes leak test, the pallet ships, the phone does not ring. When it is wrong, everyone becomes an expert on friction welding within an hour. The difference between those two days is rarely the machine. It is whether someone engineered the joint, documented the settings, and kept the process inside its window.

Spin welding rewards preparation and punishes assumption. Design the joint for the load it carries, choose parameters from diameter and material rather than habit, and treat the data log as your cheapest insurance policy.

The single takeaway: treat the weld joint as an engineered feature, not an afterthought. Do that, and a half-second of rotation quietly holds a thousand litres of chemistry where it belongs.

Frequently Asked Questions

What is spin welding of plastic in simple terms?

It is a friction welding process where one thermoplastic part is spun against a stationary part under pressure until the rubbing surfaces melt, then rotation stops and the parts are held together while the melt solidifies into a bond. It only works on joints with a circular axis, such as caps, housings, fittings and filter bodies.

Can you spin weld polyethylene and polypropylene?

Yes. PE and PP are semi-crystalline polyolefins that are generally well suited to rotational friction welding, and they are among the most common spin-welded materials. Cross-linked grades are the exception and range from difficult to impossible to weld, so always confirm the specific grade before committing to the process.

How fast is a spin welding cycle?

Most production cycles run in the range of a few seconds to roughly 20 seconds, depending on part diameter, material and cooling requirements. Friction phases commonly take about 0.5 to 8 seconds and hold/cooling commonly takes about 4 to 10 seconds, with loading and unloading added on top.

Why does spin welding produce flash and plastic particles?

Before true melting begins, softened surface material is sheared off and rolled into small particles, and once melting is established, excess melt is thrown outward as flash. Good joint design controls where that material goes using flash traps and skirts rather than trying to eliminate it.

Can spin welding create a watertight or hermetic seal?

Yes, a well-designed shear or tongue-and-groove joint can produce a repeatable hermetic seal, which is why the process is used for floats, filter housings, reservoirs and medical components. The seal depends on full circumferential melt and adequate hold pressure, so a joint that is out of round or under-clamped will leak.

What is the difference between direct-drive and inertia spin welding?

Direct-drive machines use a servo or motor to hold a programmed RPM throughout the weld and then brake, which gives fine control over speed and pressure stages. Inertia machines spin a flywheel to a set speed, then engage the parts and let stored kinetic energy do the welding, which is simple and fast but less flexible.

Does spin welding work on glass-filled plastics?

It can, but the reinforcement does not cross the weld line, so the joint does not get the same strength benefit the base resin enjoys. Glass-filled grades also tend to be more abrasive and more brittle at the interface, so expect shorter tooling life and more parameter development.

Can I spin weld two different plastics together?

Sometimes, but only when the two resins are chemically compatible and have similar melt temperatures and melt flow behaviour. Dissimilar pairs such as polycarbonate and ABS can work, while pairs with very different melting points usually produce a weak joint. Always run a trial weld and a destructive test before designing around the combination.

What causes a weak spin weld?

Common causes include too little friction energy, excessive pressure that squeezes molten material out of the joint, contamination such as mould release or oil, parts that are out of round, and releasing clamp force before the melt has solidified. Cross-sectioning a failed part usually identifies which of these is happening.